8158

• UCREFRP Catalog Number: 8158
• Author: Osmundson, D. B.
• Title: Longitudinal Variation in Fish Community Structure and Water Temperature in the Upper Colorado River; Final Report.
• USFW Year: 1999.
• USFW - Doc Type: Grand Junction, Colorado.
• Copyright Material: NO

I used lengths and weights of all captured individuals from both seasons to develop one
length-weight relationship for each species that could then be used as a standard to which
relative condition could be compared when relative condition data were partitioned by strata,
season, or life-history phase. Slope and intercept coefficients were derived by regressing log-
transformed mass as a function of length. Relative condition of each individual fish was then
calculated. A T-test was used to test for differences between mature and immature fish of
each species (within seasons), and ANOVA and the Tukey-Kramer multiple comparison test
to compare condition among seasons and among strata within seasons. To increase sample
sizes for comparisons and look for large-scale longitudinal patterns, data were combined from
adjacent strata in similar sections of the river. The resulting five groupings were strata 1-3, 4-
6, 7-9, 10-11, and 12 (the single Gunnison River stratum).
Food Habits
Stomach samples from the three major native species (bluehead sucker, flannelmouth sucker
and roundtail chub) were collected from each stratum during each season. A total of 10 fish
of each type were picked at random from each stratum; three from each of two study reaches
and four from the third reach. Stomach contents of roundtail chub were removed using a
Seaburg stomach sampler (Seaburg 1957) as modified by Gengerke et al. (1973). Back-flush
tubes of various diameters, based on the size of the fish, were used. Chubs were then released
alive. Roundtail chubs containing no food were replaced with individuals that yielded at least
some stomach contents. For suckers, individuals were sacrificed and entire digestive tracts
were removed. No field inspection was made to determine emptiness. Stomach contents of
Colorado pikeminnow captured during a concurrent study in 1994 (see Osmundson et al.
1998) were also examined using a Seaburg stomach sampler. Empty stomachs were recorded
in the field. Stomach contents were collected from Colorado pikeminnow opportunistically
from throughout the river without regard to quotas or minimum numbers per stratum.
All gastrointestinal samples were preserved in ethanol and sent to the Larval Fish Laboratory
(Colorado State University) for analysis. In the lab, extraneous tissue was dissected away
from each sucker gut and percent fullness of the gut was visually assessed. Guts that
appeared to be 100% full were used as a standard to estimate percent fullness of other guts.
To reduce laboratory time, only the anterior 1/3 of each gut was used, and sub-samples of
contents were analyzed. Food items were identified to lowest practical taxon, and visual
estimates were made of percentage of total volume of stomach contents contributed by each
taxon. The `aggregate percent method' (Swanson et al. 1974) was used to calculate an
overall percent volume for each food item for those fish containing food. Due to limited
numbers of chub and sucker stomach samples per stratum per sampling period, we combined
data from adjacent strata in similar sections of the river; using the same stratum groupings
described above for the body condition analysis. To do so, we first calculated the mean for
each stratum and then calculated the mean of these stratum means for each stratum group.
7